getting the proper transformer for a 40kw system

[winnie]

I am speaking well outside of my area of expertise here, so I'd suggest a small salt lick to go with my words.

In the situation described, using inverters which do not establish a neutral and which are not tied to ground, and with a 208V wye to 480V delta transformer, I _would_ ground the X0 of the 208V side.

While _power_ is being fed _from_ the inverters to the 480V system, the inverters themselves are taking their 'reference' _from_ the transformer. The inverter output voltage and phasing are all set by the voltage _from_ the transformer. Grounding X0 establishes the ground potential of the entire solar panel/inverter/transformer system.

Normally, if you backfeed a 208V wye to 480V delta transformer, the power source itself has a grounded neutral. In this case it is an error to ground the 208V X0 because now you have _two_ different lines in the system being grounded. When everything is working normally, this is about as bad as having multiple neutral to ground faults in a system; you just get a bit of 'objectionable current' in your EGC or GEC. But in a single phase situation, the utility supply neutral and the transformer derived neutral are trying to be at two very different voltages. You basically have a line to ground bolted fault, without OCPD on this 'line'.

To summarize: the 'no-no' is to have _two different_ neutral points grounded in a given system. Normally this means _DO NOT_ ground X0 in a reverse fed transformer. But IMHO if the only place neutral is grounded is in the 'back fed' transformer, then this is the correct thing to do.

-Jon

 

[ggunn]

I am speaking well outside of my area of expertise here, so I'd suggest a small salt lick to go with my words.

In the situation described, using inverters which do not establish a neutral and which are not tied to ground, and with a 208V wye to 480V delta transformer, I _would_ ground the X0 of the 208V side.

While _power_ is being fed _from_ the inverters to the 480V system, the inverters themselves are taking their 'reference' _from_ the transformer. The inverter output voltage and phasing are all set by the voltage _from_ the transformer. Grounding X0 establishes the ground potential of the entire solar panel/inverter/transformer system.

Normally, if you backfeed a 208V wye to 480V delta transformer, the power source itself has a grounded neutral. In this case it is an error to ground the 208V X0 because now you have _two_ different lines in the system being grounded. When everything is working normally, this is about as bad as having multiple neutral to ground faults in a system; you just get a bit of 'objectionable current' in your EGC or GEC. But in a single phase situation, the utility supply neutral and the transformer derived neutral are trying to be at two very different voltages. You basically have a line to ground bolted fault, without OCPD on this 'line'.

To summarize: the 'no-no' is to have _two different_ neutral points grounded in a given system. Normally this means _DO NOT_ ground X0 in a reverse fed transformer. But IMHO if the only place neutral is grounded is in the 'back fed' transformer, then this is the correct thing to do.

-Jon

FWIW it is not normal practice to ground the neutral (if it has one) at the inverter.

 

[alfiesauce]

further along in the study

further along in the study

Why don't you use 480V native inverters and avoid the issue and the expense of a transformer? For 40kW AC you could use a pair of Solectria or Sunny Tripower 20kW inverters.

Also, I was told by a transformer company when I had the opposite problem (480V inverters and a 208V interconnection) that transformers (theirs, at least) are not step-up to step-down interchangeable. With the same terminals they had different part numbers for up and down conversion.

The system is produced by 10K solar, a Minnesota bred company, they have not come out with a 480v inverter system yet. The plus' of the 10k system is that it works off of a low voltage dc string system, so rather than pushing 600V it's only a, don't quote me, 52V system DC. Their panel cells are also wired in a series/parallel manner. This gives them a large partial panel shade advantage, as partial shading produces minor differences in inverter voltage inputs and therefore they remain working in a more efficient manner. They also parallel their micro inverters, and turn them on sequentially, thus not making all the inverters work all of the time, saving on, in theory, inverter replacement costs.
But I digress...

The inverters I have found out, are a true 208 single phase output, they do not require a neutral.
When combined on a 3 phase panel board, they will output 208 3 phase.
So the plan, at this point, barring some futher thinking by my inspector and his boss, is that we are going to install the inverter 3phase 208 output onto the wye without an x0 point connection. Because I to was told that I cannot just arbitrarily use this step down transformer by this company for a step up transformer unless I connect with without the x0 connection. The main difference between the transformers seemed to be which side the voltage taps were on...
this is as of 4:29pm thursday afternoon

 

[ggunn]

The system is produced by 10K solar, a Minnesota bred company, they have not come out with a 480v inverter system yet. The plus' of the 10k system is that it works off of a low voltage dc string system, so rather than pushing 600V it's only a, don't quote me, 52V system DC. Their panel cells are also wired in a series/parallel manner. This gives them a large partial panel shade advantage, as partial shading produces minor differences in inverter voltage inputs and therefore they remain working in a more efficient manner. They also parallel their micro inverters, and turn them on sequentially, thus not making all the inverters work all of the time, saving on, in theory, inverter replacement costs.
But I digress...

The inverters I have found out, are a true 208 single phase output, they do not require a neutral.
When combined on a 3 phase panel board, they will output 208 3 phase.
So the plan, at this point, barring some futher thinking by my inspector and his boss, is that we are going to install the inverter 3phase 208 output onto the wye without an x0 point connection. Because I to was told that I cannot just arbitrarily use this step down transformer by this company for a step up transformer unless I connect with without the x0 connection. The main difference between the transformers seemed to be which side the voltage taps were on...
this is as of 4:29pm thursday afternoon

Ah, yes, those guys. I saw their display at SPI in Orlando year before last; it looked interesting. Good luck and let us know how it turns out.

 

[Smart $]

...

To summarize: the 'no-no' is to have _two different_ neutral points grounded in a given system. Normally this means _DO NOT_ ground X0 in a reverse fed transformer. But IMHO if the only place neutral is grounded is in the 'back fed' transformer, then this is the correct thing to do.

-Jon

FWIW it is not normal practice to ground the neutral (if it has one) at the inverter.

I agree with Jon... and that's why I was asking Bob/iwire about the source of the low voltage side of those backfed transfomers scenarios with 'melted' parts.

The reason you don't normally ground at neutral at inverters falls precisely within the concept of grounding vs. dual grounding that Jon presented.

FWIW, under the NEC, a 208Y/120 system is required to either ground X0 or the neutral at the first disconnecting means... and where that happens is also where the system GEC connection is made.

 

[Smart $]

...
So the plan, at this point, barring some futher thinking by my inspector and his boss, is that we are going to install the inverter 3phase 208 output onto the wye without an x0 point connection. Because I to was told that I cannot just arbitrarily use this step down transformer by this company for a step up transformer unless I connect with without the x0 connection. The main difference between the transformers seemed to be which side the voltage taps were on...
this is as of 4:29pm thursday afternoon

As I mentioned in my most recent reply, under the NEC, a 208Y/120 system is required to either ground X0 or the neutral at the first disconnecting means... and where that happens is also where the system GEC connection is made.

Sorry to appear redundant, but I think you and others directly involved should be aware of this, and didn't want to take the risk that you didn't read my other post.

 

[GoldDigger]

But, FWIW, if you do not connect anything to X0, do you still have a 208Y/120 system or just a 208 delta with an uncommon source?

 

[Smart $]

But, FWIW, if you do not connect anything to X0, do you still have a 208Y/120 system or just a 208 delta with an uncommon source?

I was considering that possibility when I posted (both times ), but ruled it out by recollection only. Since you brought it up, I wanted to verify before responding. The pertinent section is 250.20(B)...

(B) Alternating-Current Systems of 50 Volts to 1000
Volts. Alternating-current systems of 50 volts to 1000 volts
that supply premises wiring and premises wiring systems
shall be grounded under any of the following conditions:

(1) Where the system can be grounded so that the maximum
voltage to ground on the ungrounded conductors
does not exceed 150 volts

(2) Where the system is 3-phase, 4-wire, wye connected in
which the neutral conductor is used as a circuit conductor

(3) Where the system is 3-phase, 4-wire, delta connected in
which the midpoint of one phase winding is used as a
circuit conductor

Sure looks like it falls under both conditions (1) and (2). Although condition (2) would perhaps be negated if there wasn't the fourth wire, as long as there is an X0 terminal, condition (1) would apply.

The other possibility I was thinking of was whether the above would apply if all was a line-side PV System connection. The question I'm left with on this is whether we can include a transformer not integral with Article 690 equipment as part of the PV System. My rationale for not proceding further in this respect is because 250.20 general statement says AC systems supplying premises wiring and premises wiring systems. WHile that last part seems a bit redundant, I don't believe there's any getting around it as long as the transformer widings are configured 208Y/120V. A 208 delta configuration is the only way I see around it.

 

[jaggedben]

Sure looks like it falls under both conditions (1) and (2). Although condition (2) would perhaps be negated if there wasn't the fourth wire, as long as there is an X0 terminal, condition (1) would apply.
.

FWIW, I don't think that condition 2 applies to the OP's situation, since it's apparently not a system "in which the neutral conductor is used as a circuit conductor". In this case it seems there might not necessarily be a neutral conductor at all. But I don't think that condition 2 would apply to any system that qualifies for 705.95(B). I believe that this would exempt nearly all GT solar systems from this condition.

As for condition 1, supposing there isn't actually a neutral conductor, and the only sense in which the system 'can' be grounded is the existence of an XO terminal that would not otherwise be used and may exist only because a more appropriate piece of equipment was not available...
I would hope than an AHJ would defer to an engineer's opinion on what is safest, rather than attempt to split hairs over the meaning of the code language.

 

[Smart $]

...

As for condition 1, supposing there isn't actually a neutral conductor, and the only sense in which the system 'can' be grounded is the existence of an XO terminal that would not otherwise be used and may exist only because a more appropriate piece of equipment was not available...
I would hope than an AHJ would defer to an engineer's opinion on what is safest, rather than attempt to split hairs over the meaning of the code language.

Well the existence of an X0 terminal makes it moot to 'suppose' regarding condition (1).

As to the AHJ deferring to an engineer's opinion isn't permitted. I understand it may happen, but under the NEC, the AHJ is allowed to approve the installation, or not. They cannot rewrite Code as they see fit on a case by case basis. To do it properly, the AHJ would have to enact a Code amendment to get around this.

 

[jaggedben]

Well the existence of an X0 terminal makes it moot to 'suppose' regarding condition (1).

As to the AHJ deferring to an engineer's opinion isn't permitted. I understand it may happen, but under the NEC, the AHJ is allowed to approve the installation, or not. They cannot rewrite Code as they see fit on a case by case basis. To do it properly, the AHJ would have to enact a Code amendment to get around this.

I don't agree that the code language is clear; I think it's a matter of interpretation in any case.

Furthermore, there's 90.4.